It’s Time to Reevaluate WAR Positional Adjustments

Baseball has always been a game driven by statistics, and as the game has evolved so has those stats. Many baseball analysts think that the culmination of this statistical evolution has been the development of WAR or Wins Above Replacement. I will not go into the specifics of calculating this complicated statistic because that is not the point of this article. WAR has become the go-to statistic for baseball nerds and educated fans alike. It is an all-around statistic involving offense, defense, and base running and adjusts according to a particular player’s position. The current positional adjustments factor in defensive difficulty. Thus positions like corner outfield and first base get downgrades and positions like shortstop and catcher get upgrades. The exact breakdown of the adjustments used are as follows (from Fangraphs):

Catcher: +12.5 runs (all are per 162 defensive games)
First Base: -12.5 runs
Second Base: +2.5 runs
Third Base: +2.5 runs
Shortstop: +7.5 runs
Left Field: -7.5 runs
Center Field: +2.5 runs
Right Field: -7.5 runs
Designated Hitter: -17.5 runs

These adjustments come from comparisons involving the UZR of players who have played multiple positions. As it can be seen, designated hitters are hampered significantly by the position they play and it is one reason why David Ortiz never finds himself near the top of any WAR leaderboards. While I believe positional adjustments are necessary, I think that the defensive adjustments are unfair to certain positions, especially DHs. I think that these adjustments need to be rethought, and I think one way to reevaluate these adjustments would be to change to offensive adjustments.

Below is a chart that looks at the wOBA for each position from 2010-2014 as well as the entire league wOBA. I then took the difference for each season (wOBAleague-wOBApostition) and divided by wOBA scale which is basically the calculation for wRAA. The second to last row shows the average per plate appearance for those five seasons and then I multiplied out to calculate the positional adjustment over the last five seasons extrapolated over 600 PAs a typical season’s worth for an everyday player.

2010 wOBA0.310.3440.3160.3010.3220.3270.3260.3380.3330.326
2011 wOBA0.3090.3380.3060.3020.310.320.3210.3340.3380.321
2012 wOBA0.3120.3310.3020.30.320.3240.3230.3270.3390.32
2013 wOBA0.3070.3330.3050.2980.3140.3170.3170.3250.3340.318
2014 wOBA0.3050.3280.2980.2970.3160.3180.320.3160.3210.315
Diff 20100.013-0.0140.0080.0200.003-0.0010.000-0.010-0.006
Diff 20110.009-0.0130.0120.0150.0090.0010.000-0.010-0.013
Diff 20120.006-0.0090.0140.0160.000-0.003-0.002-0.006-0.015
Diff 20130.009-0.0120.0100.0160.0030.0010.001-0.005-0.013
Diff 20140.008-0.0100.0130.014-0.001-0.002-0.004-0.001-0.005
Runs per AB0.01- 
Runs per 6005.40-7.016.909.671.71-0.57-0.66-3.81-6.17 

So taking a look at the chart there are some major similarities and a few major differences between the offensive and defensive adjustments. The biggest positive adjustments go to shortstop, second base, and catcher, which all make sense because those are positions where offense is sacrificed for defense. One major difference is that first basemen actually have the largest negative adjustment, beating out DHs by almost one full run.

Now obviously these adjustments are not perfect, and people who prefer the defensive adjustments argue that offensive averages fluctuate too much to use them for these adjustments. I think that if stat-heads were to switch to this type of adjustment it would need to be calculated on a per year basis as opposed to the constant numbers that are used in the defensive adjustment.

Using offensive adjustments might not be the perfect way of calculating WAR, but I personally think that it is better than the current system. It rewards players like Troy Tulowitzki, Ian Desmond, and others who play defense-first positions and excel at the plate. The ideal system may be a combination of both offensive and defensive adjustments.

2 Responses

  1. TigerDoc

    WAR is a big, complicated math problem with quite a bit of subjective assumptions, not just these defensive “adjustments”. It is also has never been validated, especially regarding actual wins, which the name implies. It has some value, but it is not so valuable to have it be the one “go to” stat.

    • Paul Mammino

      There are many arguments against it but if you read up on how it is derived it is extremely useful and valuable for comparing players. For many in the baseball community it is the best statistic we currently have. The direct win correlation is not the point of it as it compares players to a “replacement level player” who is a fictional player not the player directly backing up the player. Most of the aspects of it are actually compared to league averages.


Leave a Reply